ubuntu-bugcontrol team mailing list archive

[Louis Bouchard <louis.bouchard@xxxxxxxxxxxxx>]

Hello,

Le 18/07/2013 17:22, Matthew Fischer a écrit :
> Can you explain the reasoning behind the Importance that you set these
> bugs to or whether you agree with that someone else set them to?
> 
> 
>     LVM cluster command failure :
>     https://bugs.launchpad.net/ubuntu/precise/+source/lvm2/+bug/833368
> 
>     - Helped out document and process the SRU. Reproduced the problem &
>     confirmed the fix.
>      * Assigned myself to the issue
>      * Updated description for SRU
>      * Invalidated unneeded/irrelevant tasks
> 

For this bug, importance was set by the sponsor who reworked the
proposed patch. The clustered LVM functionality was broken and did
provoke dysfunctional behaviour upon reboot : impossibility to import
clustered VG which could cause reboot issues

>     Samba Win95 printing :
>     https://bugs.launchpad.net/ubuntu/+source/samba/+bug/967410
> 
>     - Reproduced the issue, identified the upstream fix & lead the SRU
>     process to published fix.
>      * Assigned myself
>      * Adjusted Status
>      * Updated description with SRU template
> 

Here, the only place where I did change the importance was for the
Quantal task where the fix had already been implemented. For the other
tasks, the fact that it affected more than 50 people and did break an
interoperability function that worked well in the previous LTS release
explains the "High" Importance.

>     debootstrap reports corrupted files :
>     https://bugs.launchpad.net/ubuntu/+source/debootstrap/+bug/1023069
> 
>     - Reproduced the issue, identified the upstream fix & lead the SRU
>     process to published fix.
>      * Took over ownership
>      * Updated description with SRU template
>      * Understood that "Fix Released" was not when the fix had been uploaded
> 

Again here, Importance was set to "High" as the bug did break a very
basic functionality for an LTS release (unattended install using PXE).

I did not set it myself; it was done after discussion with Adam Stokes.

>     kernel thread hang on iscsi disconnect :
>     https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1056746
> 
>     - Reproduced the issue, identified the upstream fix & lead the SRU
>     process to published fix.
>      * Took over ownership
>      * Idendified upstream kernel commit fixing the issue
>      * Adapted subject to match kernel context
>      * Requested new bug tasks to someone with privs
> 

With this bug, Importance was set by kernel team members after
discussion that outlined the fact that this bug did break reboots when
iSCSI was being used. The resulting situation was a complete kernel
hang. It did definitively break accessibility of the server upon reboot.

>     dante cannot use tmpdir :
>     https://bugs.launchpad.net/duplicity/+bug/1005901
> 
>     - Reproduced the issue, identified the upstream fix & lead the SRU
>     process to published fix.
>      * Requested appropriate tasks to someone with privs
>      * Kept status in sync with bug fix
> 

For this later case, Importance was set to Medium as the impact is on a
non-core application and only happens in specific contexts. Though an
advertised functionality did not work, the context of it happening
(limited space in /tmp, specific restore mode) did not justify for  a
higher importance.

I hope that those answers are satisfactory.

Kind regards,

..Louis

-- 
Louis Bouchard
Backline Support Analyst
Canonical Ltd
Ubuntu support: http://canonical.com/support